Nova Southeastern University
NSUWorks
HCNSO Student Theses and Dissertations HCNSO Student Work
5-3-2018
Patterns in Caribbean Coral Spawning
Anna C. JordanNova Southeastern University, [email protected]
Follow this and additional works at:https://nsuworks.nova.edu/occ_stuetd
Part of theMarine Biology Commons, and theOceanography and Atmospheric Sciences and Meteorology Commons
Share Feedback About This Item
This Thesis is brought to you by the HCNSO Student Work at NSUWorks. It has been accepted for inclusion in HCNSO Student Theses and Dissertations by an authorized administrator of NSUWorks. For more information, please [email protected].
NSUWorks Citation
Anna C. Jordan. 2018. Patterns in Caribbean Coral Spawning. Master's thesis. Nova Southeastern University. Retrieved from NSUWorks, . (468)
Thesis of
Anna C. Jordan
Submitted in Partial Fulfillment of the Requirements for the Degree of
Master of Science
M.S. Marine Biology
Nova Southeastern University
Halmos College of Natural Sciences and Oceanography May 2018
Approved: Thesis Committee Major Professor: Nicole Fogarty Committee Member: Joana Figueiredo
Committee Member: Margaret Miller
HALMOS COLLEGE OF NATURAL SCIENCES AND OCEANOGRAPHY
PATTERNS IN CARIBBEAN CORAL SPAWNING
By
Anna Jordan
Submitted to the Faculty of
Halmos College of Natural Sciences and Oceanography in partial fulfillment of the requirements for the degree of Master of Science with a specialty in:
Marine Biology
Nova Southeastern University
Acknowledgements
First, I would like to thank my advisor Dr. Nicole Fogarty for her help and guidance with this project. I have gained a lot of experience being in her lab. Not only was Dr. Fogarty a fantastic teacher, but she was a great mentor as well. I was given the opportunity to help on many different projects, and I learned a great deal. I would not have been able to complete this project without Dr. Fogarty’s spawning resources, contacts, and ideas. I cannot thank her enough for all of the opportunities I was given during my time in her lab.
I would also like to thank my committee members Dr. Joana Figueiredo and Dr. Margaret Miller for their help throughout this entire process. Dr. Figueiredo was an immense help with the statistics involved with this project. I would not have been able to complete this project without her assistance. She was always willing to meet with me or refer me to statistics resources when needed. Dr. Miller also offered me valuable advice and assistance as well. Dr. Miller was able to provide me with spawning observations and is very knowledgeable about coral spawning which helped for this project. My committee members were invaluable resources for me during this process, and I am very grateful.
I need to thank my family and friends for their never ending support of me during this process. They were an immense help whether they were editing drafts for me,
offering advice, or just helping to relieve some of the stress. Their motivation and encouragement kept me going through this entire process. I could not have done this without them.
Finally, I need to thank all of the researchers who contributed spawning observations to this study. Multiple people gave me access to their published and
unpublished spawning observations in order to complete this project. My analysis would not have been as strong without these observations, and this project may not have been possible without the contribution of these scientists.
Table of Contents
LIST OF FIGURES………..i
LIST OF TABLES……….iii
ABSTRACT………...iv
Chapter 1 – Introduction ……….1
1.1 Coral Reproduction………...21.2 Factors Influencing Coral Health and Reproduction………….………..5
1.3 Objectives………..9
Chapter 2 – Publication………..11
2.1 Introduction………….………..11
2.2 Methods………..………15
2.3 Results………...………..17
2.3.1 Descriptive Statistics for Spawning………...17
2.3.2 Temporal Trends………...………34 2.3.3 Congener Trends……...………....47 2.3.4 Environmental Effects………...47 2.4 Discussion………...………49
Chapter 3 – Discussion………56
Data Limitations…..………..56 Stochastic Events…..……….………59 Congener Trends………63 Environmental Effects………64Appendix 1 – Contributors……….66
Appendix 2 – Supplemental Figures………...74
i
List of Figures
Figure 1 Geographic distribution of observations……….17
Figure 2 Spawning months by species………..18
Figure 3 Spawning days by species………..18
Figure 4 Spawning times by species……….19
Figure 5 Acropora cervicornis spawning scatterplot and moonrise bar graph…………..21
Figure 6 Acropora palmata spawning scatterplot and moonrise bar graph.………..22
Figure 7 Colpophyllia natans spawning scatterplot and moonrise bar graph….………...23
Figure 8 Dendrogyra cylindrus spawning scatterplot and moonrise bar graph.…………24
Figure 9 Diploria Labyrinthiformis spawning scatterplot and moonrise bar graph...26
Figure 10 Montastraea cavernosa spawning scatterplot and moonrise bar graph………27
Figure 11 Orbicella annularis spawning scatterplot and moonrise bar graph…...28
Figure 12 Orbicella faveolata spawning scatterplot and moonrise bar graph..………...29
Figure 13 Orbicella franksi spawning scatterplot and moonrise bar graph………...31
Figure 14 Pseudodiploria strigosa spawning scatterplot and moonrise bar graph...32
Figure 15 Stephanocoenia intersepta spawning scatterplot and moonrise bar graph...33
Figure 16 Acropora cervicornis spawning time and days by year………35
Figure 17 Acropora palmata spawning times and days by year………36
Figure 18 Colpophyllia natans spawning times and days by year……….37
Figure 19 Dendrogyra cylindrus spawning times and days by year………..38
Figure 20 Diploria labyrinthiformis spawning times and days by year………39
Figure 21 Montastraea cavernosa spawning times and days by year………...41
Figure 22 Orbicella annularis spawning times and days by year……….42
Figure 23 Orbicella faveolata spawning times and days by year………..43
Figure 24 Orbicella franksi spawning times and days by year………..44
ii
Figure 26 Stephanocoenia intersepta spawning days by year………...46
Figure 27 Acroporids Survival Analyses………...48
Figure 28 Orbicella spp. Survival Analyses………..48
Figure S1 Acropora cervicornis spawning times and days by month………...75
Figure S2 Acropora palmata spawning times and days by month………75
Figure S3 Colpophyllia natans spawning times and days by month……….76
Figure S4 Dendrogyra cylindrus spawning times and days by month………..76
Figure S5 Dendrogyra cylindrus spawning times and days by gender……….77
Figure S6 Diploria labyrinthiformis spawning times and days by month……….77
Figure S7 Montastraea cavernosa spawning days by month….……..……….78
Figure S8 Montastraea cavernosa spawning times by month………...78
Figure S9 Montastraea cavernosa spawning times and days by gender………...78
Figure S10 Orbicella annularis spawning times and days by month………79
Figure S11 Orbicella faveolata spawning times and days by month………79
Figure S12 Orbicella franksi spawning times and days by month………80
Figure S13 Pseudodiploria strigosa spawning times and days by month……….80
Figure S14 Stephanocoenia intersepta spawning days by month……….81
iii
List of Tables
Table 1 Split Spawning Observations………..20
iv
Abstract
Most corals worldwide are broadcast spawners that rely on synchronous gamete release for successful fertilization. Spawning synchrony may also decrease the
probability of heterospecific fertilization that may produce maladaptive hybrids. Despite the importance of reproductive timing, researchers have only recently begun to collect spawning data across coral species in the Caribbean, but these data remain to be
analyzed. This study investigates interannual, seasonal, and environmental patterns that may influence Caribbean scleractinian spawning times. The number of spawning observations varies widely among location and species. Most spawning observations were collected in Florida, Curaçao, and Flower Garden Banks National Marine Sanctuary. Acropora palmata, A. cervicornis, and Orbicella species were the most documented. The Orbicella spp. were very consistent for spawning day annually, while the acroporids were less reliable. However, the acroporids were more consistent for spawning time in minutes after sunset between years. Season and moon cycles were obvious proximate cues for spawning, but a strong influence from wind and tides was absent. Acropora cervicornis was the only species in this study which spawning was significantly affected by water temperature. For some scleractinians, the day of spawning was significantly affected by mass bleaching events; spawning could occur on earlier days than in previous years for up to two years after the event. This study highlights existing data gaps for Pseudodiploria clivosa, A. prolifera and Siderastrea siderea. Documenting spawning patterns is crucial to better understand the potential impacts of future threats on the already imperiled Caribbean corals at risk from reproductive failure. Keywords: Caribbean, Broadcast spawner, Orbicella, Acropora, Temporal isolation, coral, Spawning times
1 Chapter 1
Coral reefs’ structural intricacy provides significant commercial value, primarily through tourism, shoreline protection, and fishing (Hughes 1994; Park et al. 2002; Hawkins and Roberts 2004). Many coral reef organisms contain compounds that have been used in pharmaceuticals and for advancing medical research (Knowlton et al. 2010). In 2003, over half of all new medical research for cancer drugs involved marine
organisms (Cesar et al. 2003). The economic value of tropical coral reefs through
pharmaceuticals, shoreline protection, recreation, seafood, and tourism is estimated at US $797.359 billion worldwide (Cesar et al. 2003). However, human use can also cause great harm, e.g. coastal development to support tourism, interfering diving behavior, boat anchoring, and destructive or overfishing practices (Cesar et al. 2003). For decades coral reefs have been declining worldwide from these practices and other anthropogenic influences, such as disease outbreaks, pollution, and bleaching events, i.e, where corals lose their symbiotic dinoflagellate that gives the tissue its color (Hoegh-Guldberg et al. 2017; Hughes et al. 2018). These factors have led researchers to suggest that coral reefs will not survive more than a few decades without immediate protection from human exploitation (Pandolfi et al. 2003; Gattuso et al. 2015).
The Atlantic and Caribbean have 7.6% of coral cover worldwide (Spalding et al. 2001) and in 2003, the net value of coral reefs for the Caribbean and United States was almost US $80 billion (Cesar et al. 2003). Regardless of their value, Caribbean corals are particularly vulnerable to overfishing and pollution due to a lack of (or at least
insufficient) protective measures (Jackson et al. 2014). In 2012, the mean live coral cover found in the Caribbean was 16.8%, which represents almost a 20% absolute decrease since 1973 (Jackson et al. 2014). In the Florida Reef Tract, the combination of a growing human population, lower quality and quantity of fresh water input from the Everglades makes this coral reef ecosystem unique in the Western Atlantic and Caribbean, as it is one of the most studied but also heavily used reef systems in the Caribbean (Spalding et
al. 2001; Keller and Causey 2005).
Scientists are researching techniques to restore coral populations in areas that have been affected by direct and indirect anthropogenic stressors. One strategy uses the
2 natural process of fragmentation, where a piece of coral breaks off from the parent
colony, reattaches to the substrate, and continues to grow (Highsmith 1982). Scientists are using fragmentation methods to grow and transplant corals to areas where high coral mortality has occurred from disease and predator outbreaks, bleaching, or ship
groundings (Yeemin et al. 2006). Asexual reproduction is beneficial for corals because it only requires one coral, it requires less energy, is quick, and because fragments do not disperse over a large distance, the coral has a genotype adapted to the local environment (Williams 1975). However, scientists are also attempting to use sexual propagation to generate larvae and re-seed the reefs (Marhaver et al. 2015). In areas affected by multiple anthropogenic stressors, natural coral recruitment can be limited; therefore, in highly disturbed areas, restoration efforts may be the most effective and rapid way to increase coral biomass, and thus guarantee the success of sexual reproduction, and restoration of ecosystem function (Yeemin et al. 2006).
1.1 Coral Reproduction
There are two forms of sexual reproduction in corals, brooding and broadcast spawning (Marshall and Stephenson 1933; Szmant-Froelich et al. 1980). The main difference between these two reproductive strategies is that fertilization and embryonic development occur internally in brooders, and externally in broadcast spawners (Lacaze-Duthiers 1873; Marshall and Stephenson 1933; Szmant-Froelich et al. 1980; Fadlallah and Pearse 1982). The embryo develops into a planula larva that settles on the benthos, undergoes metamorphosis, and if it survives to an adequate size will reproduce,
completing the life cycle (Fadlallah 1983). In the Indo-Pacific, tens to hundreds of species can spawn in a highly synchronized mass spawning event. Yet, scientists first observed this phenomenon only a few decades ago (Babcock et al. 1986; Richmond and Hunter 1990; McGuire 1998). Worldwide most coral species (84%) are broadcast spawners (Baird et al. 2009); however, in the Caribbean broadcast spawning and
brooding species are about equally represented (Harrison and Wallace 1990). Broadcast spawners typically reproduce once or twice per year while brooders reproduce several times per year, as frequently as every month (Richmond and Hunter 1990; McGuire 1998). Regardless of their mating strategy, approximately 73% of coral species are
3 hermaphrodites, where one coral has both female and male sex organs, while other coral species have separate sexes, i.e., gonochores (Harrison and Wallace 1990; Richmond and Hunter 1990; Fine et al. 2001). Due to the exchange of genetic material between eggs and sperm from separate colonies and the recombination of their genetic material, sexual reproduction leads to increased genetic diversity (Crow 1994). Selfing, where sperm and eggs from an individual colony or clone successfully fertilize (Carlon 1999), is relatively rare, and it is unclear if the resulting larvae are viable.
Mass spawning events aid the reproductive success of the coral species that participate in this event. These advantages include increasing genetic diversity through cross fertilization of multiple synchronized genotypes and promoting higher larval survival rate due to predator satiation (Harrison et al. 1984). Fish and reef invertebrates, from brittle stars to whale sharks, will consume coral gametes and embryos until they are satiated (Westneat and Resing 1988). If corals are present in high densities and spawning is highly synchronous, predators become satiated before all coral gametes and embryos are consumed (Harrison et al. 1984). This allows the population to have higher
fertilization and larval survival. However, spawning synchrony does not come without a cost, i.e., increased chance of polyspermy and thus egg death (Styan 1998). There is also the chance of a single catastrophic event during spawning reducing reproductive success (Harrison et al. 1984; Richmond and Hunter 1990) and the potential of maladaptive or infertile hybrid formation leading to gamete wastage (Willis et al. 1997). Nevertheless, spawning is overall an effective means to reproduce.
Like many other marine species (Kojis and Quinn 1981; Caspers 1984; Hoppe and Reichert 1987; Babcock et al. 1992), for corals, a precise combination of
environmental conditions is required to induce mass spawning events (Shlesinger and Loya 1985), including temperature (van Woesik et al. 2006), light (Shlesinger and Loya 1985; Babcock et al. 1994), wind (Mangubhai and Harrison 2006; van Woesik 2009), tides (Babcock et al. 1986), genetics (Knowlton et al. 1997; Levitan et al. 2011), and chemical cues (Atkinson and Atkinson 1992; Van Veghel 1994; Slattery et al. 1999). Levitan et al. (2011) found that Orbicella spp. (formerly Montastraea) have very precise interannual spawning times, but spawning becomes less precise when corals release
4 gametes later after the full moon and sunset. If these corals do not release gametes within 15 minutes of peak spawning, fertilization is reduced (Levitan et al. 2004). In other marine invertebrates, like sea urchins, when sperm are competing, spawning precision on the scale of tens of seconds can affect which individuals mate and which do not (Levitan 2005).
Corals use seawater temperature and solar insolation cycles to synchronize
reproduction to the same season (van Woesik et al. 2006). Temperature can be influenced by sunlight and seasonal cycles and generally affects the season that corals will spawn. The monthly average sea surface temperature significantly correlates with the timing of spawning for 12 species of Caribbean broadcast spawners (van Woesik et al. 2006). Others suggest that the rate of change in sea surface temperature, not the monthly average is the proximate cue for mass spawning events (Keith et al. 2016). There are arguments that solar insolation cycles are better predictors of coral spawning in the Caribbean (van Woesik et al. 2006). Others hypothesize that spawning day is influenced by lunar factors such as the coincidence of the third quarter of the lunar cycle and the movement of the moon over the equator (Wolstenholme et al. 2018). Lin and Nozawa (2017) monitored 42 scleractinian species, including Acropora spp., at the same locations in the Indo-Pacific for seven years, allowing them to identify variability in spawning time or date that occurred between years. They found that different species follow different biological clock models. For example, acroporids are more sensitive to changes in the environment because they follow an hourglass biological clock model, which can increase the
variability in spawning time for this genus (Lin et al. 2013; Lin and Nozawa 2017).
Coral spawning has been found to coincide with calm periods in regional wind fields and low-amplitude tides. This enables the corals to have maximum fertilization success and retain larvae (van Woesik 2009). If the corals spawn during periods of high winds, their larvae could be transported to unsuitable habitats. Mangubhai and Harrison (2006) observed multiple species of corals spawning during calm periods in regional wind fields on a reef near the equator off the coast of Kenya. Tides influence the
direction that the gametes disperse and therefore impact fertilization success. Babcock et
5 The seasonal photoperiod may influence the month corals spawn (Babcock et al. 1994) and sunlight and moonlight may influence the time corals spawn (Babcock et al. 1992). Light, including moonlight and sunlight, is the most commonly recorded cue to influence coral spawning. Light cues have been found to dictate which night the corals will spawn, usually measured in the number of days before or after the full moon
(Babcock et al. 1994). In the Caribbean, broadcast spawners typically reproduce three to six days after the full moon and two to four hours after sunset. Brooding corals, such as
Porites astreoides, reproduce in relation to the new moon (Babcock et al. 1986;
Chornesky and Peters 1987). Light also has an influence on predation during coral spawning. If there is less light (i.e., before moonrise), visual predators may not be as successful at preying on coral gametes and embryos (Babcock et al. 1992).
Lastly, genetics and chemical cues appear to play a role in fine scale spawning synchrony and fertilization success of corals. Gametes could be genetically incompatible and lead to fertilization failure (Knowlton et al. 1997). Individual corals with the same genotype (ramets) tend to have similar spawning times, but unique genotypes can have significantly different spawning times (Levitan et al. 2011). Hormones also influence the timing of gametogenesis and spawning e.g., soft corals have an increase in testosterone before spawning events, and increases in progesterone were correlated with female gametogenesis (Slattery et al. 1999). Neighboring corals, regardless of genotype, spawn more synchronously than corals with the same genotype that were spaced further apart, suggesting that hormones may be a cue for mass spawning events (Levitan et al. 2011). Estradiol-17β was found to be present during mass spawning of scleractinian corals in Australia (Atkinson and Atkinson 1992). It was hypothesized that the presence of this steroid suggests that it is involved in spawning synchrony or the final maturation of the eggs (Atkinson and Atkinson 1992).
1.2 Factors influencing coral health and reproduction
Direct and indirect anthropogenic stressors are a threat to the continued existence of corals. The most prevalent threat to corals is ocean warming and acidification (Hoegh-Guldberg et al. 2017). An increase in the amount of carbon dioxide in the atmosphere causes an increase in the amount of carbon dioxide in the ocean (Raven et al. 2005). The
6 carbon dioxide in the ocean reacts with the water resulting in an increased concentration of hydrogen ions. This increase in hydrogen ions lowers the pH of the water, thus making the ocean more acidic (Raven et al. 2005). Since coral skeletons are composed/built of calcium carbonate, the increase in ocean acidity weakens their skeletons and forces corals to allocate their metabolic energy differently (Kleypas and Langdon 2006). In acidic conditions, corals need to allot more energy to build their skeletons, leading to a reduction in the energy available for other important processes, such as reproduction (Hoegh-Guldberg et al. 2007). Reduced energy allocation can cause corals to produce smaller or less viable eggs and sperm, halt reproduction in order to conserve energy, or reabsorb gametes (Szmant and Gassman 1990). The detrimental effects of ocean
acidification are found to increase in sperm-limited circumstances. Albright et al. (2010) found a compounded decrease of 73% in the number of settled larvae under ocean acidification conditions.
The increased carbon dioxide in the atmosphere can also cause warming (Callendar 1938). When heat leaves the earth’s surface and travels through the
atmosphere, the increased levels of carbon dioxide gas trap the heat in our atmosphere causing the temperature in our atmosphere to increase (Callendar 1938). Changes in temperature cause coral stress which can lead to a number of different reactions from the coral. One reaction could be that the corals may expel their endosymbionts and thus lose their color. The white calcium carbonate skeleton can be seen through their translucent tissue; therefore, this process is called “coral bleaching.” The endosymbionts provide enough nutrition through the process of photosynthesis to meet the requirements for the coral’s metabolic respiration (Muller-Parker et al. 2015). Bleaching therefore contributes to a reduction of energy reserves which will ultimately limits sexual reproduction. The reduced energy prevents or reduces the production of gametes for up to two years after the bleaching event causing a significant decrease in reproduction (Szmant and Gassman 1990; Omori et al. 2001; Levitan et al. 2014). If the corals have enough energy for gametogenesis to occur, the gametes produced could be of lesser quality, for example sperm with decreased motility (Omori et al. 2001) or eggs have less lipids (Michalek-Wagner and Willis 2001) which can reduce fertilization success and dispersal distances. Bleached corals may be able to spawn but only if they have enough energy stores (Fitt et
7
al. 2000). After a bleaching event, there can be a decrease in spawning synchrony for up
to two years (Levitan et al. 2014). Paxton et al. (2015) found that when corals are experiencing increased temperatures, even before bleaching occurs, egg volume and sperm number decrease. Elevated temperatures can also cause corals to reproduce earlier in the lunar cycle (Crowder et al. 2014; Paxton et al. 2015).
Direct anthropogenic influences, including pollution, physical contact with corals, and sedimentation can also reduce coral reproductive success. Nutrient enrichment and pollution have been found to decrease coral reproduction and growth rates (Richmond 1993). Even low levels of pollutants, for example oil, runoff, or sewage, can have a severe impact on the ecosystem over time, by causing mortality, reducing gamete production, larval recruitment, and therefore recovery (Richmond 1993). Some of the toxic substances from runoff, such as oil, have been shown to shrink the gonad size of scleractinian corals and further decrease the coral population’s ability to recover from other stressors (Rinkevich and Loya 1979). Macroalgae thrive in nutrient-rich water, including coastal environments with heavy runoff containing fertilizers from agriculture and residential lawns. This in addition to the loss of important herbivores through disease and overfishing has led to an overabundance in macroalgae (Carpenter 1990; Hughes 1994). It was also found that corals in the presence of macroalgae have lower larval output. Tanner (1995) found that corals that were cleared of macroalgae produced over twice as many larvae as the corals that were naturally overgrown. Direct diver contact causes coral injuries, potentially increasing the prevalence of disease and decreasing their reproductive potential (Lamb et al. 2014). Sedimentation from dredging and runoff also poses a threat to coral settlement, growth, and reproduction (Fabricius 2005; Fourney and Figueiredo 2017). Low light and sedimentation have been proven to reduce coral
fecundity and recruitment (Fabricius 2005). Fertilization occurs more slowly, and eggs and sperm are produced in lower quantities when adults are exposed to sedimentation (Gilmour 1999). It should be noted that when studying these stressors and their influence on a community, that there is rarely only one stressor acting on the community.
Coral diseases occur as a response to stress acting on the corals which can have a detrimental effect on coral reproduction. White-band disease is one of the most prevalent
8 coral diseases and causes more than 90% mortality in Caribbean acroporids. The disease kills coral tissue as it spreads from base to tip (Aronson and Precht 2001). Aronson and Precht (2001) found that Acropora species in the Caribbean exhibit a lack of genetic diversity due to the mortality from white-band disease. The lack of genetic diversity is then further magnified by their primarily asexual reproduction (Aronson and Precht 2001). This loss of A. palmata genetic diversity sometimes leads to the domination of one clone on a reef (Baums et al. 2006), thus reducing the ability for successful fertilization because selfing is limited (Fogarty, Vollmer, et al. 2012). Additionally, A. palmata contracts white pox disease. This disease causes circular lesions to form on the coral and eventually results in tissue loss and colony mortality. The high likelihood of mortality makes it challenging for A. palmata populations to recover from white pox outbreaks because this species relies heavily on asexual fragmentation (Patterson et al. 2002). Yellow band disease is caused by a bacterial pathogen that primarily affects Orbicella species. This disease seems to mainly affect the corals’ endosymbionts. It has been found that polyps infected with yellow band disease have significantly fewer eggs than polyps without the disease, which could decrease fertilization success (Weil et al. 2009). Lastly, Borger and Colley (2010) found that O. faveolata colonies affected by white plague disease have fewer reproductive polyps, lower reproductive mesenteries, lower oocyte volume, a lower quantity of oocytes, and lower fecundity than healthy colonies.
Fish are both predators and protectors of corals. While some fish species help the corals by eating algae, others eat coral polyps (Francini-Filho et al. 2008)and/or gametes. In Australia, Acanthochromis polyacanthus and Abudefduf bengalensis prey upon coral gametes their stomachs can be over 90% full of gametes (Westneat and Resing 1988), likely reducing coral fertilization. Parrotfish are known to selectively feed on adult corals in the Caribbean (Francini-Filho et al. 2008), particularly O. annularis polyps with greater reproductive potential, as defined by the number of gonads, number of eggs, and number of eggs per gonad between polyps (Rotjan 2007). Additionally, because
parrotfish tend to graze on the same coral polyps repeatedly, the corals need to constantly regenerate these parts likely leading to decreased reproductive rates. On the other hand, parrotfish algal grazing may have a significant positive affect on coral health by
9 disease for corals (Jackson 2001). Despite the dichotomy between the positive and
negative effects of parrotfish on corals, it has been suggested that overall parrotfish are helpful to coral health, and conservation efforts should focus on the parrotfish protection to stabilize coral populations (Mumby et al. 2007). If corals are undisturbed, they will grow faster than predators can eat them (Jackson 1977; Jackson 2001). Parrotfish have been overfished in many Caribbean locations leading to an inverse relationship between coral and algal cover (Mumby 2006).
While several studies have compiled spawning data on scleractinian species in the Caribbean and Western Atlantic, there has not been a statistical analysis of this data. Most of the studies with compiled spawning data include scleractinian species from across the world (Richmond and Hunter 1990; Baird et al. 2009; Harrison 2011). Some of these studies examine general patterns in the reproductive biology or evolution of the species (Baird et al. 2009; Harrison 2011). While the spawning of several Caribbean species has been recorded, there is no broad analysis of species-specific proximate cues for spawning. It is important to establish trends in coral reproduction prior to further coral mortality and environmental changes in order to implement the best protective measures and restoration strategies.
1.3 Objectives
The main objectives of this paper were to create a database of Caribbean, Western Atlantic, and Gulf of Mexico scleractinian spawning data, and then to identify potential proximate cues for spawning on regional scales within individual species and among congeners. This project aimed to answer three questions:
1. Is there a species-specific temporal pattern of spawning?
I analyzed annual, monthly, and daily patterns in the moon cycle and day cycle.
2. Do congeners exhibit similar spawning patterns?
I analyzed daily differences in spawning among congeners relative to moon cycle and day cycle in Acropora spp. and Orbicella spp. 3. What are the environmental proximate cues for spawning?
10 I determined if water temperature, wind speed, and moonrise time cued spawning.
11 Chapter 2
2.1 Introduction
The Atlantic Ocean and Caribbean Sea have an estimated 21,600 km2 of coral reef, equaling 7.6% of the total reef cover in the world (Spalding et al. 2001). The structural complexity of these reefs hosts immense biodiversity, providing significant commercial value through tourism, shoreline protection, and fishing (Hughes 1994; Park
et al. 2002; Hawkins and Roberts 2004). Since the 1970s, the Caribbean has decreased in
coral cover almost 20% (Côté et al. 2005; Jackson et al. 2014). Overpopulation and tourism, in combination with unenforced or absent measures of protection, overfishing, loss of herbivores, disease, and coastal pollution were found to be the main drivers behind coral habitat loss (Jackson et al. 2014). These factors led to a phase shift from a coral dominated benthos to a macroalgal domination (Côté et al. 2005). Successful coral reproduction through asexual propagation or larval reseeding coupled with increased herbivory would help to reverse this trend and restore coral reefs to their former coral dominated state.
There are two ways corals can sexually reproduce, broadcast spawning and brooding (Marshall and Stephenson 1933; Szmant-Froelich et al. 1980). The main difference between these two reproductive strategies is that fertilization and embryonic development occur internally in brooders and externally in broadcast spawners (Lacaze-Duthiers 1873; Marshall and Stephenson 1933; Szmant-Froelich et al. 1980; Fadlallah and Pearse 1982). Worldwide most coral species (84%) are broadcast spawners (Baird et
al. 2009); however, in the Caribbean broadcast spawning and brooding species are about
equally represented (Harrison and Wallace 1990). Broadcast spawners only reproduce once or twice per year, while brooders reproduce several times per year (Richmond and Hunter 1990; McGuire 1998). A majority of corals (73%) are hermaphrodites, where one coral will have both male and female sex organs, while other corals have separate sexes,
i.e., gonochores (Harrison and Wallace 1990; Richmond and Hunter 1990; Fine et al.
2001). Due to the exchange of genetic material between eggs and sperm from separate colonies and recombination, sexual reproduction leads to increased genetic diversity (Crow 1994).
12 Some species of corals are known for their precise spawning times. For example
Orbicella species (Levitan et al. 2011) have an interannual standard deviation in
spawning times of as little as 7 minutes. However, spawning becomes less precise when corals release gametes later after the full moon and sunset (Levitan et al. 2011).
Fertilization is reduced if corals do not release gametes within 15 minutes of peak spawning (Levitan et al. 2004); therefore, spawning synchrony is crucial to reproductive success. Shlesinger and Loya (1985) suggested that a precise combination of
environmental conditions is required to induce mass spawning events, including temperature (van Woesik et al. 2006), light (Shlesinger and Loya 1985; Babcock et al. 1994), wind (Mangubhai and Harrison 2006; van Woesik 2009), tides (Babcock et al. 1986), genetics (Knowlton et al. 1997; Levitan et al. 2011), and chemical cues (Atkinson and Atkinson 1992; Van Veghel 1994; Slattery et al. 1999).
For many broadcast spawning corals, the monthly average sea surface
temperature, which is influenced by sunlight and seasonal cycles, is the best predictor of which month corals will spawn (van Woesik et al. 2006). Yet, it has been found that the rate of change in sea surface temperature, not the monthly average, is the proximate cue for mass spawning events (Keith et al. 2016). However, solar insolation cycles have been suggested to be better predictors of coral spawning in the Caribbean (van Woesik et al. 2006). Solar insolation is the quantity of electromagnetic energy incident to earth’s surface (van Woesik et al. 2006). Others hypothesize that spawning day is influenced by lunar factors such as the coincidence of the third quarter of the lunar cycle and the movement of the moon over the equator (Wolstenholme et al. 2018). Multiple coral species spawn during calm periods in regional wind fields, and mass spawning frequently coincides with low-amplitude tides (Babcock et al. 1986; Mangubhai and Harrison 2006). Spawning under both of these conditions enables the corals to have maximum
fertilization success and retain larvae (van Woesik 2009). Lunar light cues dictate which night corals will spawn and also influences predation (Babcock et al. 1994). A majority of spawning observations also occur before the moon rises to reduce predation on coral gametes and embryos during mass spawning events (Babcock et al. 1994). Spawning for most species of scleractinians in the Caribbean occurs after sunset, the exception being
13 cues appear to play a role in fine scale spawning synchrony and fertilization success of corals. Individual corals with the same genotype (ramets) have similar spawning times, but unique genotypes have significantly different spawning times (Levitan et al. 2011). Hormones also influence the timing of gametogenesis and spawning (Slattery et al. 1999). Neighboring corals, regardless of genotype, spawn more synchronously than corals with the same genotype that were spaced further apart, suggesting that hormones may be a cue for mass spawning events (Levitan et al. 2011).
There are many environmental and anthropogenic factors that threaten coral reproductive success; among the most detrimental is thermal stress. Coral bleaching often occurs under thermal stress and causes the coral to expel its endosymbionts
(Symbiodinium spp.) and lose its color. These endosymbionts produce most of the coral’s nutrition, and their loss can have devastating effects on the coral (Muller-Parker et al. 2015). If the coral does not have sufficient metabolic energy, they will reallocate their energy in order to survive. The reduced energy from bleaching can prevent or reduce production of gametes for up to two years after the bleaching event, causing a substantial decrease in reproductive output (Szmant and Gassman 1990; Omori et al. 2001; Levitan
et al. 2014). If the corals have enough energy for gametogenesis to occur, the gametes
produced could be of lesser quality, for example fewer sperm with decreased motility (Omori et al. 2001) and reduced egg volume including decreased lipids in eggs
(Michalek-Wagner and Willis 2001; Paxton et al. 2015). Broadcast spawners reproduce earlier under increased sea surface temperature. Furthermore, after a bleaching event, there can be a decrease in spawning synchrony for up to two years (Levitan et al. 2014; Paxton et al. 2015). Brooders are not immune to the effects of thermal stress. Elevated temperatures likely caused Pocillopora damicornis to release planulae earlier in the lunar cycle, and the shifts can occur quickly, in as little as one reproductive cycle (Crowder et
al. 2014; Levitan et al. 2014; Paxton et al. 2015).
Disease outbreaks and predation have detrimental effects on coral reproduction. White plague disease and yellow band disease reduce the number of reproductive polyps in Orbicella species (Weil et al. 2009; Borger and Colley 2010). Polyps infected with yellow band disease have significantly less eggs than polyps without the disease (Weil et
14
al. 2009). Orbicella (formerly Montastraea) faveolata colonies infected with white
plague disease had lower oocyte volume, less oocytes, and lower fecundity than healthy colonies (Borger and Colley 2010). Parrotfish have been found to selectively graze on corals in the Caribbean (Francini-Filho et al. 2008), specifically grazing on O. annularis polyps with higher reproductive potential, defined as number of gonads and number of eggs (Rotjan 2007). Acanthochromis polyacanthus and Abudefduf bengalensis have been found to eat until their stomachs are over 90% full of gametes (Westneat and Resing 1988), likely reducing coral fertilization. This is why coral mass spawning events usually occur after sunset and before moonrise; the darkness reduces predation on coral gametes and embryos (Babcock et al. 1994).
To implement the best protective measures and restoration strategies, it is important to establish trends in coral reproduction prior to further coral mortality and environmental changes. Several studies have compiled spawning data on scleractinian species, but an emphasis on the Caribbean is lacking. Most of the studies with compiled spawning data include scleractinian species from across the world (Richmond and Hunter 1990; Baird et al. 2009; Harrison 2011). Some of these studies examine general patterns in the reproductive biology or evolution of the species (Baird et al. 2009; Harrison 2011). While compiled spawning data exists for some Caribbean species, no broad analysis of spawning data or the environmental factors that may influence spawning exists for Caribbean scleractinians.
The main objectives of this study were to create a database of Caribbean, Western Atlantic, and Gulf of Mexico scleractinian spawning information to determine which spawning cues and environmental factors influence spawning within a species and among congeners. Specifically, I tested whether water temperature, wind speed, or time of moonrise affected the absence or presence of spawning.
15 2.2 Methods
Data was compiled from peer-reviewed publications, NOAA’s Coral Health and Monitoring Program Coral ListServer, posts to the coral spawning research Facebook page (created and managed by N. Fogarty), and contributing researchers (see Appendix 1) across the Western Atlantic Ocean, Caribbean Sea, and Gulf of Mexico for every species of scleractinian coral for which data was available. The information collected from these sources included date and time of spawning, proportion of corals observed that spawned, and environmental data at the time of the spawning event, including sea surface temperature, time of moonrise, wind speed, and tides. If environmental data was not provided, it was obtained from other sources. Sunset and moonrise data was gathered from the United States Naval Observatory Astronomical Applications Department website, http://aa.usno.navy.mil/data/docs/RS_OneDay.php. Temperature data was obtained from Rutgers University Coastal Ocean Observation Lab website,
https://marine.rutgers.edu/cool/sat_data/?product=sst®ion=floridacoast¬humbs=0. Wind data was obtained from the Weather Underground website,
https://www.wunderground.com. Tide data was sourced from NOAA’s Tides and Currents database https://tidesandcurrents.noaa.gov/historic_tide_tables.html.
Data Analysis
Statistical tests were chosen by the distribution of data being tested to explore patterns that might exist with Western Atlantic scleractinian coral spawning times. To assess species-specific temporal patterns, I used Mantel-Haenszel survival analyses and Mann-Whitney Wilcoxon tests. The Mantel-Haenszel tests were run for each individual species that had at least 20 spawning observations and across genera. Survival analyses were used to compare day cycle, defined as spawning minutes after sunset, or moon cycle, defined as spawning days after the full moon, by month and by year. I also tested if there were patterns found between males and females for spawning day or spawning time for the gonochoric species in this study. I used a survival analysis because it tests the time leading up to an event (i.e., spawning), which does not have to be death. The test
analyzes the time in minutes or days leading up to spawning. Each individual curve represented a different month or year, in order to compare trends between these factors.
16 Data quality control was performed before tests were run, which involved removing invalid data points. The Mann-Whitney Wilcoxon test was used to analyze whether bleaching events affected spawning day for individual species of corals. The year in which a mass bleaching event occurred (i.e., 1998, 2005, and 2010) and the two years following the bleaching year were tested against other years to determine if the spawning day was significantly different during this time.
To determine whether congeners exhibit similar spawning patterns, I used Mantel-Haenszel survival analyses as well. These tests analyzed moon cycles and day cycles for the acroporid and Orbicella spp. congeners. For these survival analyses, each individual curve represented a different congener.Data quality control was performed before these tests were run also.
To analyze the environmental proximate cues for spawning, I used Generalized Linear Models (GLMs). These were run on adequate environmental data and adequate negative spawning observations, i.e., a minimum of 20 observations with all variables present, to test whether water temperature, wind speed, or moonrise time affected the presence or absence of spawning. The species tested were A. cervicornis, A. palmata, O.
17 2.3 Results
2.3.1 Descriptive Statistics for Spawning
The data spanned 26 different regions across the Caribbean, Gulf of Mexico, and Western Atlantic (Fig. 1). The top three regions with the most spawning observations in the dataset were Curaçao, the Florida Keys, and Flower Garden Banks National Marine Sanctuary. There was a sufficient amount of data to run analyses on 11 species of broadcast spawning scleractinians. Of these species, three were gonochores and eight were hermaphrodites (Table S1). For some species, data goes back to 1983, while the most recent observations are from 2016. Spawning month refers to the month that spawning occurs. In this dataset spawning occurs from May to November, with all 11 species having August as a peak spawning month (Fig. 2, Table S1). Diploria
labyrinthiformis had the broadest spawning window from May to November (Table S1).
Spawning day refers to the day spawning occurred and is recorded as days before or after the full moon. Spawning day ranged from the day of the full moon to 18 days after the full moon for the species in this study (Fig. 3). Spawning time refers to the precise time
171 285 323 <10 observations 10-50 observations 51-100 observations 101-150 observations
150 or more observations (top 3)
Figure 1. Geographic distribution of observations. The dots represent regions where observations were recorded.
18 (hours and minutes) spawning occurred, and it was recorded as minutes before or after sunset or moonrise. Most species spawned between 30 and 300 minutes after sunset; exceptions were D. labyrinthiformis and Montastraea cavernosa because they also spawned before sunset (Fig. 4). Seven species displayed “split-spawn”, i.e. spawned after two consecutive full moons within the same region (Table 1).
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 A. cervicornis A. palmata A. prolifera O. annularis O. faveolata O. franksi C. natans D. labyrinthiformis P. strigosa D. cylindrus M. cavernosa S. intersepta
May June July August September October November
A. cervicornis A. palmata A. prolifera O. annularis O. faveolata O. franksi C. natans D. labyrinthiformis P. strigosa D. cylindrus M. cavernosa S. intersepta
Figure 2. Spawning months by species. Gray represents a month in which spawning occurs. Black represents the peak spawning month(s).
Figure 3. Spawning days by species. The numbers correspond to the date after the full moon with day 0 referring to the date of the full moon. Gray represents a day in which spawning occurs. Black represents the peak spawning days. Peak spawning day was not included for A.
19
Acropora cervicornis spawning occurred from July to September, peaking in
August, and varied greatly across spawning day and time (Figs. 5a, S1). Spawning varied across days after the full moon for all three months where spawning was observed (Fig. 5a). July and August had more variability in spawning day than September (Fig. 5a). Days 3, 5, and 6 after the full moon had more variance in spawning time than other days where spawning was observed. Observations for A. cervicornis spawning ranged from 30 to 257 minutes after sunset and 1 to 15 days after the full moon (Figs. 5a, S1). Spawning peaked between 150 and 165 minutes after sunset on days 3 to 6 after the full moon. The variability in spawning time changed from day to day (Fig. 5a). Most A. cervicornis spawning (75%) was observed prior to 30 minutes post-moonrise. The variability, however, was high, ranging from 536 minutes before moonrise to 307 minutes after moonrise (Fig. 5b). One split spawn was observed for A. cervicornis, in July and August 1985 (Table 1).
Acropora palmata spawned from July to September and had high variability for
both spawning day and spawning time for all three months (Figs. 6a, Table S1). Peak
-12 0 -> -91 -90 -> -61 -60 -> -31 -30 -> -1 0 -29 30-59 60-89 90 -119 120 -149 150 -179 180 -209 210 -239 240 -269 270 -299 300 -329 A. cervicornis A. palmata A. prolifera O. annularis O. faveolata O. franksi C. natans D. labyrinthiformis P. strigosa D. cylindrus M. cavernosa S. intersepta
Figure 4. Spawning times by species. The numbers correspond to the minutes after sunset that spawning occurred with negative numbers referring to the minutes before sunset. Gray represents that spawning was observed within that 30 minute time frame.
20 spawning for A. palmata occurred in August. July had less variability in spawning time than August or September (Fig. 6a). Days 3 and 4 after the full moon had more variance
Species Site Month Year Days After Full Moon
Full Moon Date
Acropora cervicornis
La Parguera, Puerto Rico July August
1985 7,8 7/2/85 7/31/85 Key Largo, FL August 1997 13
6,8
7/19/97 8/18/97 Tres Palmas, Puerto Rico August 2007 4
3
7/30/07 8/28/07 Acropora palmata Elbow Reef, Florida
Keys August September 2012 2 2,3 8/2/12 8/31/12 La Bocana Chica, Mexico July August 2013 4 3 7/22/13 8/20/13 Carrie Bow, Belize July
August 2013 5,6,7 2,3 7/22/13 8/20/13 Montastraea cavernosa
Flower Garden Banks August September 1995 6,7,12 8 8/10/95 9/8/95 Orbicella annularis
Key Largo, FL August 1997 13 6,8
7/19/97 8/18/97 Seaquarium, Curaçao September
October
2015 6 8/29/15
9/27/15 Flower Garden Banks August
September
1995 7,8 8
8/10/95 9/8/95 Flat Cay, USVI St.
Thomas August September 2012 8 7 8/2/12 8/31/12 Orbicella faveolata
Hind Bank, USVI St. Thomas August September 2012 8 7 8/2/12 8/31/12 Horseshoe Reef, Florida
Keys August September 2014 6,7 6 8/10/14 9/9/14 Seaquarium, Curaçao September
October
2015 6 8/29/15
9/27/15 Orbicella franksi Key Largo, FL August 1997 13
6,8
7/19/97 8/18/97 Flower Garden Banks August
September 2001 5,6,7 7,8 8/4/01 9/2/01 Pseudodiploria strigosa
Flower Garden Banks August September
1995 7,8,11 8
8/10/95 9/8/95 Table 1. Summary of observed split spawns for 7 species across Florida, the Gulf of Mexico, and the Caribbean from data collected for this study.
21 in spawning time than other days where spawning was observed. Acropora palmata had the greatest range of observed spawning days of any species in this study, with
observations recorded from the day of the full moon to 18 days after the full moon (Figs. 6a, S2a). Spawning time also varied from 50 to 260 minutes after sunset (Figs. 6a, S2b). Peak spawning occurred 3 to 5 days after the full moon from 136 to 157 minutes after sunset, but had the highest variability in spawning times for A. palmata (Figs. 6a). There was also high variability in spawning in relation to moonrise (Fig. 6b). Spawning had been observed starting at 611 minutes before moonrise until 138 minutes after moonrise
25 75 125 175 225 275 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Min ut es After Su nset
Days After Full Moon
July August September
0 2 4 6 8 10 12 -54 0 to -511 -51 0 to -481 -48 0 to -451 -24 0 to -211 -21 0 to -181 -18 0 to -151 -15 0 to -121 -12 0 to -91 -90 to -6 1 -60 to -3 1 -30 to -1 0 t o 29 30 to 5 9 60 to 8 9 90 to 1 19 120 t o 1 49 150 t o 1 79 300 t o 3 29 Freq uen cy
Minutes After Moonrise
// //
Figure 5. Acropora cervicornis spawning in Western Atlantic. a) Average spawning times for a specific site, error bars denote range of spawning time for that evening. b) Number of spawning observations for 30 min time block after moonrise (negative values indicate spawning before moonrise).
a
22 (Fig. 6b). Three-quarters of spawning observations occurred prior to 7 minutes after moonrise (Fig. 6b). There were five split spawns observed for A. palmata (Table 1). These occurred in 1997, 2007, 2012, and at two locations in 2013 (Table 1).
Colpophyllia natans was observed spawning from August to November and with
limited variance in spawning days and times (Fig. 7a, Table S1). August had the highest variability in spawning time, but the most data points (Figs. 7a, S3b). Spawning was observed from 38 to 170 minutes after sunset and 6 to 10 days after the full moon (Figs.
50 70 90 110 130 150 170 190 210 230 250 270 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Min ut es After Su nset
Days After Full Moon
July August September
0 5 10 15 20 25 30 35 -62 0 to -591 -47 0 to -441 -44 0 to -411 -41 0 to -381 -32 0 to -291 -29 0 to -261 -26 0 to -231 -23 0 to -20 1 -20 0 to -171 -17 0 to -141 -14 0 to -111 -11 0 to -81 -80 to -5 1 -50 to -2 1 -20 to 9 10 to 3 9 40 to 6 9 70 to 9 9 100 t o 1 29 130 t o 1 59 Freq uen cy
Minutes After Moonrise
// //
Figure 6. Acropora palmata spawning in Western Atlantic. a) Average spawning times for a specific site, error bars denote range of spawning time for that evening. b) Number of spawning observations for 30 min time block after moonrise (negative values indicate spawning before moonrise).
a
23 7a, S3, Table S1). Peak spawning was on days 8 and 9 after the full moon from 83 to 123 minutes after sunset (Figs. 7a, Table S1). Colpophyllia natans was one of two species in this study to have all of their spawning observations occur before moonrise (Fig. 7b) and a majority (75%) of the observations occurred more than 287 minutes before moonrise (Fig. 7b). Split spawning was not recorded for C. natans.
Dendrogyra cylindrus is a gonochore and was observed spawning from August to
October, with the majority of the observations in August and September (Fig. 8a, Table S1). The spawning days and times varied the least of the 11 species studied here (Fig. S4). Male D. cylindrus were observed spawning from 2 to 5 days after the full moon, and
25 45 65 85 105 125 145 165 185 205 225 5 6 7 8 9 10 Min ut es After Su nset
Days After Full Moon
August September October November
Figure 7. Colpophyllia natans spawning in Western Atlantic. a) Average spawning times for a specific site, error bars denote range of spawning time for that evening. b) Number of spawning
observations for 30 min time block after moonrise. 0 1 2 3 4 5 6 7 8 470 t o 4 41 440 t o 4 11 410 t o 3 81 380 t o 3 51 350 t o 3 21 320 t o 2 91 290 t o 2 61 260 t o 2 31 230 t o 2 01 200 t o 1 71 80 to 5 1 Freq uen cy
Minutes Before Moonrise
// b
24 females were observed spawning from 1 to 5 days after the full moon (Figs. 8a, S5a, Table S1). Peak spawning days occurred for males from 2 to 4 days after the full moon and females from 2 to 3 days after the full moon. Males spawned from 58 to 134 minutes after sunset while females spawned from 58 to 142 minutes after sunset (Fig. 8a, Table
50 75 100 125 150 175 200 0 1 2 3 4 5 Min ut es After Su nset
Days After Full Moon
August September October
Figure 8. Dendrogyra cylindrus spawning in Western Atlantic. a) Average spawning times for a specific site, error bars denote range of spawning time for that evening. b) Number of spawning observations for 30 min time block after moonrise (negative values indicate spawning before moonrise). 0 2 4 6 8 10 12 14 16 -14 0 to -111 -11 0 to -81 -80 to -5 1 -50 to -2 1 -20 to 9 10 to 3 9 40 to 6 9 70 to 9 9 Freq uen cy
Minutes After Moonrise
b a
25 S1). Peak spawning time for males was from 93 to 119 minutes after sunset and females from 102 to 134 minutes after sunset. Dendrogyra cylindrus was the only gonochore in this study with significantly different spawning times between males and females (Mantel-Haenszel test p = 0.0451; Fig. S5b). Spawning in relation to moonrise for D.
cylindrus occurred from 132 minutes before moonrise to 73 minutes after moonrise (Fig.
8b). Three-quarters of spawning observations occurred prior to 5 minutes before moonrise (Fig. 8b). No split spawn was recorded for D. cylindrus.
Diploria labyrinthiformis had one of the smallest ranges of spawning times and
days in this study, but the largest range of spawning months (Figs. 2, 9a, S6, Table S1). This was also the only species of scleractinian in this study to spawn exclusively before sunset. Diploria labyrinthiformis spawned from May to September, with peak spawning occurring in June and August (Fig. 2). All of the months seemed to have similar
variability in spawning days and times (Figs. 9a, S6). Spawning occurred from 7 to 13 days after the full moon and 117 minutes before sunset until the time of sunset (Fig. 9a, Table S1). Day 13 after the full moon had the most variability for spawning time, while the other spawning days had consistent spawning times (Fig. 9a). Peak spawning for D.
labyrinthiformis occurred on days 11 and 12 after the full moon from 52 to 40 minutes
before sunset (Fig. 9a, Table S1). Due to the early spawning times, all colonies spawned before moonrise (Fig. 9b). Spawning was observed from 677 to 140 minutes before moonrise, with three-quarters of the observations occurring prior to 544 minutes before moonrise (Fig. 9b). No split spawning observations were recorded for D.
labyrinthiformis.
Gonochore, Montastraea cavernosa, demonstrated high variance in spawning months, days, and times. Spawning was observed from June to November, with most of the observations occurring in August and September (Figs. 2, 10a, Table S1). All of the months of spawning had high variability in spawning times and days (Figs. 10a, S7, S8). There were observations from day 1 to 12 after the full moon for both males and females (Figs. 10a, S9a, Table S1). Days 4 and 9 after the full moon had less variability in
spawning times than the other spawning days (Fig. 10a). Days 6 and 7 after the full moon were the peak spawning days for both males and females. This was the only species in
26 this study that had spawning times both before and after sunset. Males spawned from 19 minutes before sunset to 259 minutes after sunset, while females spawned from 9 minutes before sunset to 245 minutes after sunset (Fig. 10a, Table S1). Males had peak spawning from 62 to 154 minutes after sunset, while for females spawning peaked from 62 to 147 minutes after sunset (Fig. 10a, Table S1). Males and females did not have significantly different spawning times (Mantel-Haenszel test p = 0.824; Fig. S9b). There were spawning observations from 349 minutes before moonrise to 75 minutes after moonrise
0 10 20 30 40 50 60 9 10 11 12 13 Min ut es Before Su nset
Days After Full Moon
May June July August September
110 120
Figure 9. Diploria labyrinthiformis spawning in Western Atlantic. a)
Average spawning times for a specific site, error bars denote range of spawning time for that evening. b) Number of spawning observations for 30 min time block after moonrise.
0 1 2 3 4 680 t o 6 51 650 t o 6 21 620 t o 5 91 590 t o 5 61 560 t o 5 31 530 t o 5 01 140 t o 1 11 Freq uen cy
Minutes Before Moonrise
// b
a )
27 (Fig. 10b). A majority (75%) of the spawning observations occurred prior to 91 minutes before moonrise (Fig. 10b). There was one observation of split spawning for M.
-50 0 50 100 150 200 250 300 3 4 5 6 7 8 9 Min ut es After Su nset
Days After Full Moon
August September October November
Figure 10. Montastraea cavernosa spawning in Western Atlantic. a) Average spawning times for a specific site, error bars denote range of spawning time for that evening. b) Number of spawning observations for 30 min time block after moonrise (negative values indicate spawning before moonrise).
0 2 4 6 8 10 12 14 -35 0 to -321 -32 0 to -291 -29 0 to -261 -26 0 to -231 -23 0 to -201 -20 0 to -171 -17 0 to -141 -14 0 to -111 -11 0 to -81 -80 to -5 1 -50 to -2 1 -20 to 9 10 to 3 9 40 to 6 9 70 to 9 9 Freq uen cy
Minutes After Moonrise
b a
28
cavernosa in the dataset, occurring in August and September of 1995 (Table 1).
The spawning months for O. annularis across the Caribbean were August to November (Fig. 11a, Table S1). Spawning peaked in September. November was the most consistent for spawning month out of all the months (Fig. 11a). October and November had more consistent spawning times than August or September (Figs. 11a, S10b).
75 125 175 225 275 325 4 5 6 7 8 9 Min ut es After Su nset
Days After Full Moon
August September October November
Figure 11. Orbicella annularis spawning in Western Atlantic. a) Average spawning times for a specific site, error bars denote range of spawning time for that evening. b) Number of spawning observations for 30 min time block after moonrise (negative values indicate spawning before moonrise).
0 5 10 15 20 -51 0 to -481 -33 0 to -301 -30 0 to -271 -27 0 to -241 -24 0 to -211 -21 0 to -181 -18 0 to -151 -15 0 to -121 -12 0 to -91 -90 to -6 1 -60 to -3 1 -30 to -1 0 t o 29 30 to 5 9 630 t o 6 59 Freq uen cy
Minutes After Moonrise
// //
a
29 Spawning was observed from 4 to 13 days after the full moon and 93 to 308 minutes after sunset (Figs. 11a, S10, Table S1). Peak spawning occurred on days 6 and 7 after the full moon from 180 to 220 minutes after sunset (Fig. 11a, Table S1). All of the days appeared to have similar variability in spawning times from the data used for this study (Fig. 11a).
Orbicella annularis had a wide range of spawning times in relation to moonrise time, 75 100 125 150 175 200 225 250 275 300 4 5 6 7 8 9 Min ut es After Su nset
Days After Full Moon
August September October November
Figure 12. Orbicella faveolata spawning in Western Atlantic. a) Average spawning times for a specific site, error bars denote range of spawning time for that evening. b) Number of spawning observations for 30 min time block after moonrise (negative values indicate spawning before moonrise).
0 2 4 6 8 10 12 14 16 -22 0 to -201 -20 0 to -181 -18 0 to -161 -16 0 to -141 -14 0 to -121 -12 0 to -101 -10 0 to -81 -80 to -6 1 -60 to -4 1 -40 to -2 1 -20 to -1 0 t o 19 20 to 3 9 40 to 5 9 60 to 7 9 80 to 9 9 Freq uen cy
Minutes After Moonrise
b a b
30 with observations recorded from 508 minutes before moonrise until 639 minutes after moonrise (Fig. 11b). Three-quarters of the spawning observations occurred prior to 72 minutes before moonrise (Fig. 11b). Split spawning was recorded twice for O. annularis (Table 1). These occurred in August of 1997 and September and October of 2015.
Orbicella faveolata had similar spawning days, times, and months to its congener, O. annularis. August to November were the spawning months for O. faveolata, with
spawning peaking in September (Fig. 12a, Table S1). All of the spawning months seemed to have variability in their spawning times and days, with September having the most variability in spawning times (Figs. 12a, S11). Spawning occurred from day 4 to 9 after the full moon from 88 to 275 minutes after sunset (Fig. 12a, Table S1). The days with the least variability in spawning time were days 5 and 9 after the full moon (Fig. 12a). The peak spawning time was observed from 181 to 223 minutes after sunset on days 6 and 7 after the full moon, similar to O. annularis (Fig. 12a, Table S1). Orbicella faveolata had a smaller range of spawning times in relation to moonrise than O. annularis (Fig. 12b). The spawning observations ranged from 220 minutes before moonrise to 81 minutes after moonrise (Fig. 12b). Three-quarters of the spawning observations happened prior to 44 minutes before moonrise (Fig. 12b). Orbicella faveolata, along with A. palmata, had the highest number of split spawns found in the dataset (Table 1). These occurred in August and September of 1995, at two locations in August and September of 2012, August and September of 2014, and September and October of 2015 (Table 1).
While Orbicella franksi had similar spawning days and months to its congeners, it had significantly earlier spawning times than the other congeners (Fig. 13a, Table S1). The observed spawning days were similar to O. annularis, occurring from day 4 to 13 after the full moon with the peak spawning days being 6 to 8 (Figs. 13a, S12a, Table S1). All of the spawning days had high variability for spawning time with day 7 having the most variability (Fig. 13a). Orbicella franksi was observed spawning from 44 to 265 minutes after sunset, slightly earlier than its congeners, with the peak occurring from 109 to 159 minutes after sunset (Figs. 13a, S12b, Table S1). Spawning in relation to moonrise occurred from 379 minutes before moonrise to 123 minutes after moonrise, with
three-31 quarters of the observations occurring prior to 114 minutes before moonrise (Fig. 13b).
Orbicella franksi had two split spawning observations recorded in the dataset, one in
August of 1997 and one in August and September of 2001 (Table 1).
50 100 150 200 250 300 4 5 6 7 8 9 10 Min ut es After Su nset
Days After Full Moon
August September October November
Figure 13. Orbicella franksi spawning in Western Atlantic. a) Average spawning times for a specific site, error bars denote range of spawning time for that evening. b) Number of spawning observations for 30 min time block after moonrise (negative values indicate spawning before moonrise).
0 2 4 6 8 10 12 14 16 18 -38 0 to -351 -35 0 to -321 -32 0 to -291 -29 0 to -261 -26 0 to -231 -23 0 to -201 -20 0 to -171 -17 0 to -141 -14 0 to -111 -11 0 to -81 -80 to -5 1 -50 to -2 1 -20 to 9 10 to 3 9 100 t o 1 29 Freq uen cy
Minutes After Moonrise
// a
32
Pseudodiploria strigosa is a hermaphroditic broadcast spawner and had a wide
range of spawning times and days. Spawning was recorded from July to October (Fig. 14a, Table S1). Peak spawning occurred in August and September. Spawning
observations were recorded for 5 to 14 days after the full moon from 37 to 313 minutes after sunset (Figs. 14a, S13, Table S1). Peak spawning occurred on days 6 through 8 after the full moon from 108 to 173 minutes after sunset (Fig. 14a, Table S1). Spawning occurred from 395 minutes before moonrise to 47 minutes after moonrise (Fig. 14b). A
0 50 100 150 200 250 300 350 4 5 6 7 8 9 10 11 Min ut es After Su nset
Days After Full Moon
July August September October
Figure 14. Pseudodiploria strigosa spawning in Western Atlantic. a) Average spawning times for a specific site, error bars denote range of spawning time for that evening. b) Number of spawning observations for 30 min time block after moonrise (negative values indicate spawning before moonrise).
0 1 2 3 4 5 6 7 8 9 -40 0 to -371 -37 0 to -341 -34 0 to -311 -31 0 to -281 -28 0 to -251 -25 0 to -221 -22 0 to -191 -19 0 to -161 -16 0 to -131 -13 0 to -101 -10 0 to -71 -70 to -4 1 -40 to -1 1 -10 to 19 20 to 4 9 Freq uen cy
Minutes After Moonrise
a
33 majority (75%) of the spawning observations occurred prior to 81 minutes before
moonrise (Fig. 14b). There is one split spawning record for P. strigosa in August and September of 1995 (Table 1). 0 50 100 150 200 250 300 1 2 3 4 5 6 7 8 9 Min ut es After Su nset
Days After Full Moon
August September October
Figure 15. Stephanocoenia intersepta spawning in Western Atlantic. a) Average spawning times for a specific site, error bars denote range of spawning time for that evening. b) Number of spawning observations for 30 min time block after moonrise (negative values indicate spawning before moonrise). 0 1 2 3 4 -31 0 to -291 -29 0 to -271 -27 0 to -251 -25 0 to -231 -23 0 to -211 -21 0 to -191 -19 0 to -171 -17 0 to -151 -15 0 to -131 -13 0 to -111 -11 0 to -91 -90 to -7 1 -70 to -5 1 -50 to -3 1 -30 to -1 1 -10 to 9 Freq uen cy
Minutes After Moonrise
b a
